Kunal, I guess that RDF/XML and N3 should demonstrate similar speed of text parsing, in kilobytes of source text per second.
If data are the same then the count of triple insertions multiplied by average insertion time is same for both cases and the difference is obviously in some other place. TURTLE and the like are parsed by flex+bison code, RDF/XML parser is written by hands. Nevertheless both are quite fast and random variations in process/thread scheduling may provide bigger difference than serialization format. Best Regards, Ivan Mikhailov, OpenLink Software. On Wed, 2008-03-12 at 12:39 -0700, Kunal Patel wrote: > Hi, > > I have got the same RDF data in two formats, the RDF/XML > serialization format and the NTriples format. I tried loading the > RDF/XML serialization format using the DB.DBA.RDF_LOAD_RDFXML_MT and > the NTriples format using the DB.DBA.TTLP_MT. > > I was hoping that the NTriples format would load much faster then > the RDF/XML serialization format, but on the contrary the RDF/XML > serialization format loaded faster than the NTriples format. Is that > the expected behavior, or am I missing any thing. > > Kunal > > P.S. I am using the NTriples format not the N3 format or the turtle > format. > > > ______________________________________________________________________ > Looking for last minute shopping deals? Find them fast with Yahoo! > Search.
